This document discusses the development of an artificial neural network constitutive model for aluminum 7075 alloy. It begins with an introduction to constitutive equations and the typical parametric approach to constitutive modeling. It then discusses the benefits of a non-parametric artificial neural network approach. The document outlines experimental data collected on aluminum 7075 alloy under different strain rates and temperatures. It aims to design and train a neural network model to accurately model the stress-strain behavior of aluminum 7075 alloy across a range of loading conditions.
IDENTIFICATION OF DELAMINATION SIZE AND LOCATION OF COMPOSITE LAMINATE FROM TIME DOMAIN DATA OF MAGNETOSTRICTIVE SENSOR AND ACTUATOR USING ARTIFICIAL NEURAL NETWORK.
Three stress analysis methodologies were used to analyze stresses in a mild steel specimen with an eccentric hole under tension: theoretical analysis using equations, computational analysis using FEA software, and experimental analysis using strain gauges. Each method agreed the maximum stress occurred in the hole area, with the second highest in the net area and lowest in the gross area. Theoretical and experimental results differed by an average of 10%, theoretical and computational by 5.1%, and computational and experimental by 4.8%. Retesting revealed a bending moment induced by the testing machine, requiring averaging of results. Overall the different methodologies correlated well.
The document discusses two methods for mesh refinement - the p-method and h-method. The p-method increases the order of the polynomial used in the finite element model, allowing for more accurate results without changing the mesh. The h-method reduces the size of elements to create a finer mesh, better approximating the real solution in areas of high stress gradients. Both methods aim to improve the accuracy of finite element analysis results, with the p-method doing so without requiring changes to the mesh.
IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com
Isoparametric formulation allows for more accurate modeling of curved boundaries by mapping regular element shapes from a natural coordinate system to the actual curved geometric shapes in the global coordinate system. This mapping technique revolutionized finite element analysis by reducing unnecessary stress concentrations compared to using only straight-edged elements. The document goes on to define isoparametric, superparametric, and subparametric elements, and explains the basic theorems and uniqueness conditions for valid mappings between coordinate systems. It also describes Gaussian numerical integration for assembling stiffness matrices in isoparametric finite element analysis and provides some illustrative numerical examples.
Relevance Vector Machines for Earthquake Response Spectra drboon
This study uses Relevance Vector Machine (RVM) regression to develop a probabilistic model for the average horizontal component of 5%-damped earthquake response spectra. Unlike conventional models, the proposed approach does not require a functional form, and constructs the model based on a set predictive variables and a set of representative ground motion records. The RVM uses Bayesian inference to determine the confidence intervals, instead of estimating them from the mean squared errors on the training set. An example application using three predictive variables (magnitude, distance and fault mechanism) is presented for sites with shear wave velocities ranging from 450 m/s to 900 m/s. The predictions from the proposed model are compared to an existing parametric model. The results demonstrate the validity of the proposed model, and suggest that it can be used as an alternative to the conventional ground motion models. Future studies will investigate the effect of additional predictive variables on the predictive performance of the model.
The document provides an introduction to the Finite Element Method (FEM). It discusses the history and development of FEM from the 1950s to the present. It outlines the basic concepts of FEM including discretization of the domain into finite elements connected at nodes, and the approximation of displacements within each element. The document also discusses minimum potential energy theory, which is the variational principle that FEM is based on. Example problems and a tutorial are mentioned. Advantages of FEM include its ability to model complex geometries and loading, while disadvantages include increased computational time and memory requirements compared to other methods.
IDENTIFICATION OF DELAMINATION SIZE AND LOCATION OF COMPOSITE LAMINATE FROM TIME DOMAIN DATA OF MAGNETOSTRICTIVE SENSOR AND ACTUATOR USING ARTIFICIAL NEURAL NETWORK.
Three stress analysis methodologies were used to analyze stresses in a mild steel specimen with an eccentric hole under tension: theoretical analysis using equations, computational analysis using FEA software, and experimental analysis using strain gauges. Each method agreed the maximum stress occurred in the hole area, with the second highest in the net area and lowest in the gross area. Theoretical and experimental results differed by an average of 10%, theoretical and computational by 5.1%, and computational and experimental by 4.8%. Retesting revealed a bending moment induced by the testing machine, requiring averaging of results. Overall the different methodologies correlated well.
The document discusses two methods for mesh refinement - the p-method and h-method. The p-method increases the order of the polynomial used in the finite element model, allowing for more accurate results without changing the mesh. The h-method reduces the size of elements to create a finer mesh, better approximating the real solution in areas of high stress gradients. Both methods aim to improve the accuracy of finite element analysis results, with the p-method doing so without requiring changes to the mesh.
IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com
Isoparametric formulation allows for more accurate modeling of curved boundaries by mapping regular element shapes from a natural coordinate system to the actual curved geometric shapes in the global coordinate system. This mapping technique revolutionized finite element analysis by reducing unnecessary stress concentrations compared to using only straight-edged elements. The document goes on to define isoparametric, superparametric, and subparametric elements, and explains the basic theorems and uniqueness conditions for valid mappings between coordinate systems. It also describes Gaussian numerical integration for assembling stiffness matrices in isoparametric finite element analysis and provides some illustrative numerical examples.
Relevance Vector Machines for Earthquake Response Spectra drboon
This study uses Relevance Vector Machine (RVM) regression to develop a probabilistic model for the average horizontal component of 5%-damped earthquake response spectra. Unlike conventional models, the proposed approach does not require a functional form, and constructs the model based on a set predictive variables and a set of representative ground motion records. The RVM uses Bayesian inference to determine the confidence intervals, instead of estimating them from the mean squared errors on the training set. An example application using three predictive variables (magnitude, distance and fault mechanism) is presented for sites with shear wave velocities ranging from 450 m/s to 900 m/s. The predictions from the proposed model are compared to an existing parametric model. The results demonstrate the validity of the proposed model, and suggest that it can be used as an alternative to the conventional ground motion models. Future studies will investigate the effect of additional predictive variables on the predictive performance of the model.
The document provides an introduction to the Finite Element Method (FEM). It discusses the history and development of FEM from the 1950s to the present. It outlines the basic concepts of FEM including discretization of the domain into finite elements connected at nodes, and the approximation of displacements within each element. The document also discusses minimum potential energy theory, which is the variational principle that FEM is based on. Example problems and a tutorial are mentioned. Advantages of FEM include its ability to model complex geometries and loading, while disadvantages include increased computational time and memory requirements compared to other methods.
This document describes a Meshless Local Petrov-Galerkin (MLPG) Mixed Collocation Method for solving Cauchy inverse heat transfer problems. The method independently interpolates temperature and heat flux fields using the same meshless basis functions. Balance and compatibility equations are enforced through collocation at nodes. Numerical examples demonstrate the accuracy, convergence, and stability of the method for inverse heat transfer problems where both temperature and heat flux are prescribed on part of the boundary.
CONSISTENT AND LUMPED MASS MATRICES IN DYNAMICS AND THEIR IMPACT ON FINITE EL...IAEME Publication
There are two strategies in the finite element analysis of dynamic problems related to natural frequency determination viz. the consistent / coupled mass matrix and the lumped mass matrix. Correct determination of natural frequencies is extremely important and forms the basis of any further NVH (Noise vibration and harshness) calculations and Impact or crash analysis. It has been thought by the finite element community that the consistent mass matrix should not be used as it leads to a higher computational cost and this opinion has been prevalent since 1970. We are of the opinion that in today’s age where computers have become so fast we can use the consistent mass matrix on relatively coarse meshes with an advantage for better accuracy rather than going for finer meshes and lumped mass matrix
The document discusses using machine learning to develop density functional approximations for orbital-free density functional theory calculations. Specifically, kernel ridge regression is used to approximate the kinetic energy of non-interacting fermions confined to a 1D box as a functional of electron density. This machine-learned density functional approximation achieves highly accurate energies and self-consistent densities, outperforming traditional approximations. Various kernels, cross-validation methods, and representations of the electron density are explored to optimize the machine-learned approximation.
A nonlinear model for the vibration suppression of a smart composite elastic plate using graphical representation involving fuzzy control is presented. The plate follows the von Kármán and Kirchhoff plate bending theories and the oscillations are caused by external transversal loading forces, which are applied directly on it. Two different control forces, one continuous and one located at discrete points, are considered. The mechanical model is spatially discretized by using the time spectral Galerkin and collocation methods. The aim is to suppress vibrations through a simulation process within a modern graphical computing environment. Here we use MATLAB/SIMULINK, while other similar packages can be used as well. The nonlinear controller is designed, based on an application of a Mamdani-type fuzzy inference system. A computational algorithm, proposed and tested here is not only effective but robust as well. Furthermore, all elements of the study can be replaced or extended, due to the flexibility of the used SIMULINK environment.
The document discusses various numerical methods for analyzing mechanical components under applied loads, including the finite element method. It describes weighted residual methods like the Galerkin method and collocation method which approximate solutions by minimizing residuals. The variational or Rayleigh-Ritz method selects displacement fields to minimize total potential energy. The finite element method divides a structure into small elements and applies these methods to obtain approximate solutions for displacements and stresses at discrete points.
This document discusses a method for predicting the dynamic response and flutter characteristics of structures using experimental modal parameters when the exact system properties like mass and stiffness are unknown. The method uses modal parameters obtained from ground vibration tests in finite element and computational fluid dynamics software to analyze transient response and flutter speeds. It was validated on a tapered aluminum plate structure by comparing results obtained using experimental modal data to those from a finite element model using the actual material properties. Close agreement was observed between the two methods, showing this approach can accurately analyze structures without prior knowledge of system configurations.
1) The document discusses the basics of the finite element method (FEM), which involves dividing a structure into simple subdomains called finite elements connected at nodes.
2) FEM allows for the analysis of complex problems by replacing differential equations with algebraic equations at nodes. This is done using shape functions to interpolate values within an element.
3) The document compares FEM to other numerical methods like the finite difference method, noting advantages of FEM include better accuracy with fewer elements and the ability to model curved boundaries and nonlinear problems.
The document provides an introduction to the finite element method (FEM). It discusses that FEM is a numerical technique used to approximate solutions to boundary value problems defined by partial differential equations. It can handle complex geometries, loadings, and material properties that have no analytical solution. The document outlines the historical development of FEM and describes different numerical methods like the finite difference method, variational method, and weighted residual methods that FEM evolved from. It also discusses key concepts in FEM like discretization into elements, node points, and interpolation functions.
Dynamic Structural Optimization with Frequency ConstraintsIJMER
This document summarizes a research paper on dynamic structural optimization with frequency constraints. It discusses using finite element modeling and an evolution optimization technique to optimize a structure's natural frequencies. The technique calculates sensitivity parameters to determine how removing individual elements affects natural frequencies. It can increase or decrease frequencies, or change the gap between frequencies. Examples optimize a cantilever plate model to increase the first frequency, decrease the first frequency, and increase the gap between the first two frequencies. The technique provides an easy way to control dynamic characteristics and obtain new optimized topology designs using common finite element software.
The document summarizes a CFD analysis of fluid flow around a classic Rolls-Royce vehicle. The initial model showed recirculating and reversed flow behind the vehicle. To improve the design, the author modeled an airfoil-shaped rear end. Simulation results for the original and modified designs showed that the modified design with smooth edges eliminated wake and low pressure regions behind the vehicle, indicating an efficient design. The author recommends preventing circulating velocity fields and low pressure spaces behind vehicles in CFD design work.
Structural morphology optimization by evolutionary proceduresStroNGER2012
The paper deals with the identification of optimal structural morphologies through evolutionary procedures.
Two main approaches are considered. The first one simulates the Biological Growth (BG) of natural structures like the bones and the trees. The second one, called Evolutionary Structural Optimization (ESO), removes material at low stress level. Optimal configurations are addressed by proper optimality indexes and by a monitoring of the structural response. Design graphs suitable to this purpose are introduced and employed in the optimization of a pylon carrying a suspended roof and of a bridge under multiple loads.
Modal Analysis of Single Rectangular Cantilever Plate by Mathematically, FEA ...IRJET Journal
1. The document analyzes the natural frequencies and mode shapes of a single rectangular cantilever plate through mathematical, finite element analysis (FEA), and experimental methods.
2. The natural frequencies were first calculated mathematically using Euler-Bernoulli beam theory. The plate was then modeled and analyzed in ANSYS to obtain natural frequencies and mode shapes via FEA. Experimental testing was also conducted to measure the natural frequencies.
3. The results obtained from the three methods showed good correlation with each other, though some methods produced results with up to 20% error compared to others. Analyzing the plate through different techniques helped validate the results and understand the dynamic characteristics of the structure.
The document discusses modifications to the PC algorithm for constraint-based causal structure learning that remove its order-dependence, which can lead to highly variable results in high-dimensional settings; the modified algorithms are order-independent while maintaining consistency under the same conditions, and simulations and analysis of yeast gene expression data show they improve performance over the original PC algorithm in high-dimensional settings.
Identification of Material Parameters of Pultruded FRP Composite Plates using...Subhajit Mondal
This document summarizes research that used finite element model updating to identify material parameters of pultruded fiber reinforced plastic (FRP) composite plates. Researchers conducted experimental modal testing on an FRP plate to measure its vibration responses. They then used an inverse eigensensitivity method to update a finite element model of the plate, adjusting the model's in-plane elastic parameters until its responses matched experimental data. This allowed them to estimate the plate's actual elastic properties nondestructively. The technique was validated using independent quasi-static tests, demonstrating its potential for structural health monitoring of pultruded FRP infrastructure.
IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com
Finite element analysis in orthodontics/ /certified fixed orthodontic courses...Indian dental academy
The Indian Dental Academy is the Leader in continuing dental education , training dentists in all aspects of dentistry and offering a wide range of dental certified courses in different formats.
Interfacial delamination failure in bonded concrete overlay systems a review...IAEME Publication
This document reviews theories and modeling methods for describing delamination failure at the interface between two bonded cementitious materials. It discusses traditional stress-based and energy-based failure criteria approaches. It presents the interface cohesive zone model (ICZM) as a viable approach for describing and predicting delamination in bonded concrete overlay systems. The ICZM treats delamination as a progressive failure involving both crack initiation and propagation. It considers distinct analytical cases involving material and structural property variables. The concluding model shows that numerical values of delamination coefficients and energy release rates vary depending on overlay scale, problem type, and material property mismatches.
Algorithm for Modeling Unconventional Machine Tool Machining Parameters using...IDES Editor
Unconventional machining process finds a lot of
application in aerospace and precision industries. It is
preferred over other conventional methods because of the
advent of composite and high strength to weight ratio
materials, complex parts and also because of its high accuracy
and precision. Usually in unconventional machine tools, trial
and error method is used to fix the values of process
parameters. In the proposed work an algorithm which is
developed using Artificial Neural Network (ANN) is proposed
to create mathematical model functionally relating process
parameters and operating parameters of any unconventional
machine tool. This is accomplished by training a feed forward
network with back propagation learning algorithm. The
required data which are used for training and testing the ANN
in the case study is obtained by conducting trial runs in EBW
machine. By adopting the proposed algorithm there will be a
reduction in production time and set-up time along with
reduction in manufacturing cost in unconventional machining
processes. This in general increases the overall productivity.
The programs for training and testing the neural network are
developed, using MATLAB package
POWER SYSTEM PROBLEMS IN TEACHING CONTROL THEORY ON SIMULINKijctcm
This experiment demonstrates to engineering students that control system and power system theory are not orthogonal, but highly interrelated. It introduces a real-world power system problem to enhance time domain State Space Modelling (SSM) skills of students. It also shows how power quality is affected with real-world scenarios. Power system was modeled in State Space by following its circuit topology in a bottom-up fashion. At two different time instances of the power generator sinusoidal wave, the transmission line was switched on. Fourier transform was used to analyze resulting line currents. It validated the harmonic components, as expected, from power system theory. Students understood the effects of switching transients at various times on supply voltage sinusoid within control theory and learned time domain analysis. They were surveyed to gauge their perception of the project. Results from a before/after assessment analyzed usingT-Tests showed a statistically significant enhanced learning in SSM.
Power System Problems in Teaching Control Theory on Simulinkijctcm
This experiment demonstrates to engineering students that control system and power system theory are not orthogonal, but highly interrelated. It introduces a real-world power system problem to enhance time domain State Space Modelling (SSM) skills of students. It also shows how power quality is affected with real-world scenarios. Power system was modeled in State Space by following its circuit topology in a bottom-up fashion. At two different time instances of the power generator sinusoidal wave, the transmission line was switched on. Fourier transform was used to analyze resulting line currents. It validated the harmonic components, as expected, from power system theory. Students understood the effects of switching transients at various times on supply voltage sinusoid within control theory and learned time domain analysis. They were surveyed to gauge their perception of the project. Results from a before/after assessment analyzed usingT-Tests showed a statistically significant enhanced learning in SSM.
Oscillatory Stability Prediction Using PSO Based Synchronizing and Damping To...journalBEEI
This paper presents the assessment of stability domains for the angle stability condition of the power system using Particle Swarm Optimization (PSO) technique. An efficient optimization method using PSO for synchronizing torque coefficients Ksand damping torque coefficients Kd to identify the angle stability condition on multi-machine system. In order to accelerate the determination of angle stability, PSO is proposed to be implemented in this study. The application of the proposed algorithm has been justified as the most accurate with lower computation time as compared to other optimization techniques such as Evolutionary Programming (EP) and Artificial Immune System (AIS). Validation with respect to eigenvalues determination, Least Square (LS) method and minimum damping ratio ξmin confirmed that the proposed technique is feasible to solve the angle stability problems.
This document describes a Meshless Local Petrov-Galerkin (MLPG) Mixed Collocation Method for solving Cauchy inverse heat transfer problems. The method independently interpolates temperature and heat flux fields using the same meshless basis functions. Balance and compatibility equations are enforced through collocation at nodes. Numerical examples demonstrate the accuracy, convergence, and stability of the method for inverse heat transfer problems where both temperature and heat flux are prescribed on part of the boundary.
CONSISTENT AND LUMPED MASS MATRICES IN DYNAMICS AND THEIR IMPACT ON FINITE EL...IAEME Publication
There are two strategies in the finite element analysis of dynamic problems related to natural frequency determination viz. the consistent / coupled mass matrix and the lumped mass matrix. Correct determination of natural frequencies is extremely important and forms the basis of any further NVH (Noise vibration and harshness) calculations and Impact or crash analysis. It has been thought by the finite element community that the consistent mass matrix should not be used as it leads to a higher computational cost and this opinion has been prevalent since 1970. We are of the opinion that in today’s age where computers have become so fast we can use the consistent mass matrix on relatively coarse meshes with an advantage for better accuracy rather than going for finer meshes and lumped mass matrix
The document discusses using machine learning to develop density functional approximations for orbital-free density functional theory calculations. Specifically, kernel ridge regression is used to approximate the kinetic energy of non-interacting fermions confined to a 1D box as a functional of electron density. This machine-learned density functional approximation achieves highly accurate energies and self-consistent densities, outperforming traditional approximations. Various kernels, cross-validation methods, and representations of the electron density are explored to optimize the machine-learned approximation.
A nonlinear model for the vibration suppression of a smart composite elastic plate using graphical representation involving fuzzy control is presented. The plate follows the von Kármán and Kirchhoff plate bending theories and the oscillations are caused by external transversal loading forces, which are applied directly on it. Two different control forces, one continuous and one located at discrete points, are considered. The mechanical model is spatially discretized by using the time spectral Galerkin and collocation methods. The aim is to suppress vibrations through a simulation process within a modern graphical computing environment. Here we use MATLAB/SIMULINK, while other similar packages can be used as well. The nonlinear controller is designed, based on an application of a Mamdani-type fuzzy inference system. A computational algorithm, proposed and tested here is not only effective but robust as well. Furthermore, all elements of the study can be replaced or extended, due to the flexibility of the used SIMULINK environment.
The document discusses various numerical methods for analyzing mechanical components under applied loads, including the finite element method. It describes weighted residual methods like the Galerkin method and collocation method which approximate solutions by minimizing residuals. The variational or Rayleigh-Ritz method selects displacement fields to minimize total potential energy. The finite element method divides a structure into small elements and applies these methods to obtain approximate solutions for displacements and stresses at discrete points.
This document discusses a method for predicting the dynamic response and flutter characteristics of structures using experimental modal parameters when the exact system properties like mass and stiffness are unknown. The method uses modal parameters obtained from ground vibration tests in finite element and computational fluid dynamics software to analyze transient response and flutter speeds. It was validated on a tapered aluminum plate structure by comparing results obtained using experimental modal data to those from a finite element model using the actual material properties. Close agreement was observed between the two methods, showing this approach can accurately analyze structures without prior knowledge of system configurations.
1) The document discusses the basics of the finite element method (FEM), which involves dividing a structure into simple subdomains called finite elements connected at nodes.
2) FEM allows for the analysis of complex problems by replacing differential equations with algebraic equations at nodes. This is done using shape functions to interpolate values within an element.
3) The document compares FEM to other numerical methods like the finite difference method, noting advantages of FEM include better accuracy with fewer elements and the ability to model curved boundaries and nonlinear problems.
The document provides an introduction to the finite element method (FEM). It discusses that FEM is a numerical technique used to approximate solutions to boundary value problems defined by partial differential equations. It can handle complex geometries, loadings, and material properties that have no analytical solution. The document outlines the historical development of FEM and describes different numerical methods like the finite difference method, variational method, and weighted residual methods that FEM evolved from. It also discusses key concepts in FEM like discretization into elements, node points, and interpolation functions.
Dynamic Structural Optimization with Frequency ConstraintsIJMER
This document summarizes a research paper on dynamic structural optimization with frequency constraints. It discusses using finite element modeling and an evolution optimization technique to optimize a structure's natural frequencies. The technique calculates sensitivity parameters to determine how removing individual elements affects natural frequencies. It can increase or decrease frequencies, or change the gap between frequencies. Examples optimize a cantilever plate model to increase the first frequency, decrease the first frequency, and increase the gap between the first two frequencies. The technique provides an easy way to control dynamic characteristics and obtain new optimized topology designs using common finite element software.
The document summarizes a CFD analysis of fluid flow around a classic Rolls-Royce vehicle. The initial model showed recirculating and reversed flow behind the vehicle. To improve the design, the author modeled an airfoil-shaped rear end. Simulation results for the original and modified designs showed that the modified design with smooth edges eliminated wake and low pressure regions behind the vehicle, indicating an efficient design. The author recommends preventing circulating velocity fields and low pressure spaces behind vehicles in CFD design work.
Structural morphology optimization by evolutionary proceduresStroNGER2012
The paper deals with the identification of optimal structural morphologies through evolutionary procedures.
Two main approaches are considered. The first one simulates the Biological Growth (BG) of natural structures like the bones and the trees. The second one, called Evolutionary Structural Optimization (ESO), removes material at low stress level. Optimal configurations are addressed by proper optimality indexes and by a monitoring of the structural response. Design graphs suitable to this purpose are introduced and employed in the optimization of a pylon carrying a suspended roof and of a bridge under multiple loads.
Modal Analysis of Single Rectangular Cantilever Plate by Mathematically, FEA ...IRJET Journal
1. The document analyzes the natural frequencies and mode shapes of a single rectangular cantilever plate through mathematical, finite element analysis (FEA), and experimental methods.
2. The natural frequencies were first calculated mathematically using Euler-Bernoulli beam theory. The plate was then modeled and analyzed in ANSYS to obtain natural frequencies and mode shapes via FEA. Experimental testing was also conducted to measure the natural frequencies.
3. The results obtained from the three methods showed good correlation with each other, though some methods produced results with up to 20% error compared to others. Analyzing the plate through different techniques helped validate the results and understand the dynamic characteristics of the structure.
The document discusses modifications to the PC algorithm for constraint-based causal structure learning that remove its order-dependence, which can lead to highly variable results in high-dimensional settings; the modified algorithms are order-independent while maintaining consistency under the same conditions, and simulations and analysis of yeast gene expression data show they improve performance over the original PC algorithm in high-dimensional settings.
Identification of Material Parameters of Pultruded FRP Composite Plates using...Subhajit Mondal
This document summarizes research that used finite element model updating to identify material parameters of pultruded fiber reinforced plastic (FRP) composite plates. Researchers conducted experimental modal testing on an FRP plate to measure its vibration responses. They then used an inverse eigensensitivity method to update a finite element model of the plate, adjusting the model's in-plane elastic parameters until its responses matched experimental data. This allowed them to estimate the plate's actual elastic properties nondestructively. The technique was validated using independent quasi-static tests, demonstrating its potential for structural health monitoring of pultruded FRP infrastructure.
IJERA (International journal of Engineering Research and Applications) is International online, ... peer reviewed journal. For more detail or submit your article, please visit www.ijera.com
Finite element analysis in orthodontics/ /certified fixed orthodontic courses...Indian dental academy
The Indian Dental Academy is the Leader in continuing dental education , training dentists in all aspects of dentistry and offering a wide range of dental certified courses in different formats.
Interfacial delamination failure in bonded concrete overlay systems a review...IAEME Publication
This document reviews theories and modeling methods for describing delamination failure at the interface between two bonded cementitious materials. It discusses traditional stress-based and energy-based failure criteria approaches. It presents the interface cohesive zone model (ICZM) as a viable approach for describing and predicting delamination in bonded concrete overlay systems. The ICZM treats delamination as a progressive failure involving both crack initiation and propagation. It considers distinct analytical cases involving material and structural property variables. The concluding model shows that numerical values of delamination coefficients and energy release rates vary depending on overlay scale, problem type, and material property mismatches.
Algorithm for Modeling Unconventional Machine Tool Machining Parameters using...IDES Editor
Unconventional machining process finds a lot of
application in aerospace and precision industries. It is
preferred over other conventional methods because of the
advent of composite and high strength to weight ratio
materials, complex parts and also because of its high accuracy
and precision. Usually in unconventional machine tools, trial
and error method is used to fix the values of process
parameters. In the proposed work an algorithm which is
developed using Artificial Neural Network (ANN) is proposed
to create mathematical model functionally relating process
parameters and operating parameters of any unconventional
machine tool. This is accomplished by training a feed forward
network with back propagation learning algorithm. The
required data which are used for training and testing the ANN
in the case study is obtained by conducting trial runs in EBW
machine. By adopting the proposed algorithm there will be a
reduction in production time and set-up time along with
reduction in manufacturing cost in unconventional machining
processes. This in general increases the overall productivity.
The programs for training and testing the neural network are
developed, using MATLAB package
POWER SYSTEM PROBLEMS IN TEACHING CONTROL THEORY ON SIMULINKijctcm
This experiment demonstrates to engineering students that control system and power system theory are not orthogonal, but highly interrelated. It introduces a real-world power system problem to enhance time domain State Space Modelling (SSM) skills of students. It also shows how power quality is affected with real-world scenarios. Power system was modeled in State Space by following its circuit topology in a bottom-up fashion. At two different time instances of the power generator sinusoidal wave, the transmission line was switched on. Fourier transform was used to analyze resulting line currents. It validated the harmonic components, as expected, from power system theory. Students understood the effects of switching transients at various times on supply voltage sinusoid within control theory and learned time domain analysis. They were surveyed to gauge their perception of the project. Results from a before/after assessment analyzed usingT-Tests showed a statistically significant enhanced learning in SSM.
Power System Problems in Teaching Control Theory on Simulinkijctcm
This experiment demonstrates to engineering students that control system and power system theory are not orthogonal, but highly interrelated. It introduces a real-world power system problem to enhance time domain State Space Modelling (SSM) skills of students. It also shows how power quality is affected with real-world scenarios. Power system was modeled in State Space by following its circuit topology in a bottom-up fashion. At two different time instances of the power generator sinusoidal wave, the transmission line was switched on. Fourier transform was used to analyze resulting line currents. It validated the harmonic components, as expected, from power system theory. Students understood the effects of switching transients at various times on supply voltage sinusoid within control theory and learned time domain analysis. They were surveyed to gauge their perception of the project. Results from a before/after assessment analyzed usingT-Tests showed a statistically significant enhanced learning in SSM.
Oscillatory Stability Prediction Using PSO Based Synchronizing and Damping To...journalBEEI
This paper presents the assessment of stability domains for the angle stability condition of the power system using Particle Swarm Optimization (PSO) technique. An efficient optimization method using PSO for synchronizing torque coefficients Ksand damping torque coefficients Kd to identify the angle stability condition on multi-machine system. In order to accelerate the determination of angle stability, PSO is proposed to be implemented in this study. The application of the proposed algorithm has been justified as the most accurate with lower computation time as compared to other optimization techniques such as Evolutionary Programming (EP) and Artificial Immune System (AIS). Validation with respect to eigenvalues determination, Least Square (LS) method and minimum damping ratio ξmin confirmed that the proposed technique is feasible to solve the angle stability problems.
Moodle generalised formulation of laminate theory using beam fe for delaminat...Shree Bineet Kumar Kavi
This document presents a generalized formulation of laminate theory using beam finite elements to model delaminated composite beams with piezoelectric actuators and sensors. A coupled linear layerwise laminate theory and beam finite element model are developed that include delaminations as additional degrees of freedom. Numerical results illustrate the effects of delamination on sensor voltage, displacements, stresses, and modal frequencies/shapes. The model can predict the response of laminated composite beams with piezoelectric layers and various delamination sizes.
In the present paper the experimental study of
Nanotechnology involves high cost for Lab set-up and the
experimentation processes were also slow. Attempt has also
been made to discuss the contributions towards the societal
change in the present convergence of Nano-systems and
information technologies. one cannot rely on experimental
nanotechnology alone. As such, the Computer- simulations and
modeling are one of the foundations of computational
nanotechnology. The computer modeling and simulations
were also referred as computational experimentations. The
accuracy of such Computational nano-technology based
experiment generally depends on the accuracy of the following
things: Intermolecular interaction, Numerical models and
Simulation schemes used. The essence of nanotechnology is
therefore size and control because of the diversity of
applications the plural term nanotechnology is preferred by
some nevertheless they all share the common feature of control
at the nanometer scale the latter focusing on the observation
and study of phenomena at the nanometer scale. In this paper,
a brief study of Computer-Simulation techniques as well as
some Experimental result
This document discusses modeling flexible structures with bonded piezoelectric films. It begins by describing piezoelectric film properties and models. Shaped piezoelectric films are then introduced, which can have varying poling directions and etched electrode shapes. Next, equations are derived to model the interaction between flexible beams and multiple bonded piezoelectric films, accounting for film shapes. These interaction equations are then used to develop state-space models of flexible structures with piezoelectric films for control system design. Experimental validation is also discussed. The equations and models allow for active vibration damping of flexible structures using piezoelectric films.
A robust data treatment approach for fuel cells system analysisISA Interchange
The document describes a robust approach for analyzing data from fuel cell stack testing. It addresses challenges in handling large amounts of data from multiple devices. The approach includes:
1) Developing an interface in Excel to automate data handling and analysis across Excel and MATLAB for improved efficiency.
2) Using dynamic time warping to synchronize data sequences and align them based on time for better comparison.
3) Applying data reconciliation to optimally adjust measurements so they obey physical constraints like conservation of voltage sums, improving accuracy of analysis.
RESEARCH ON THE POSSIBILITIES OF USING LINEAR OBSERVATION MODELS IN WELDING P...Kiogyf
RESEARCH ON THE POSSIBILITIES OF USING LINEAR OBSERVATION MODELS IN WELDING PROCESSES
Abstract. The usage relevance of linear observation models of welding processes is substantiated. A brief analytical review of scientific papers on using linear observation models for researching the processes of arc welding is introduced. An example of studies utilizing a linear observation model based on the results of experiments with alloying of a weld metal is given. The type of linear observation model – fractional replication – is determined. An experiment planning matrix and test results are given. After performing tests and processing the experimental data, mathematical models in the form of dependencies of the transfer of alloying elements into the weld metal were obtained. According to the results of the experiments, empirical models in the form of dependencies evaluating the percentages of manganese, silicon and carbon, transfer coefficients from the “wire-to-product” distance, the diameter of the welding nozzle and the filling wire feed rate were obtained. Since the degree of uncertainty of parameters during welding is high enough, the necessity of further modification of linear observation models using expert knowledge is justified. The mathematical specification of the linear observation model with parameters in the form of fuzzy intervals is given. An algorithm for identifying the parameters of a linear observation model in the form of fuzzy intervals has been developed.
Keywords. Metal welding, observation model, regression, shielding gas, plasma-arc welding, weld bead alloying, prediction, control
Cone Crusher Model Identification Using Block-Oriented Systems with Orthonorm...ijctcm
In this paper, block-oriented systems with linear parts based on Laguerre functions is used to approximation of a cone crusher dynamics. Adaptive recursive least squares algorithm is used to identification of Laguerre model. Various structures of Hammerstein, Wiener, Hammerstein-Wiener models are tested and the MATLAB simulation results are compared. The mean square error is used for models validation.It has been found that Hammerstein-Wiener with orthonormal basis functions improves the quality of approximation plant dynamics. The mean square error for this model is 11% on average throughout the considered range of the external disturbances amplitude. The analysis also showed that Wiener model cannot provide sufficient approximation accuracy of the cone crusher dynamics. During the process it is unstable due to the high sensitivity to disturbances on the output.The Hammerstein-Wiener model will be used to the design nonlinear model predictive control application
In this paper, block-oriented systems with linear parts based on Laguerre functions is used to
approximation of a cone crusher dynamics. Adaptive recursive least squares algorithm is used to
identification of Laguerre model. Various structures of Hammerstein, Wiener, Hammerstein-Wiener models
are tested and the MATLAB simulation results are compared. The mean square error is used for models
validation.It has been found that Hammerstein-Wiener with orthonormal basis functions improves the
quality of approximation plant dynamics. The mean square error for this model is 11% on average
throughout the considered range of the external disturbances amplitude. The analysis also showed that
Wiener model cannot provide sufficient approximation accuracy of the cone crusher dynamics. During the
process it is unstable due to the high sensitivity to disturbances on the output.The Hammerstein-Wiener
model will be used to the design nonlinear model predictive control application.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Analytical model for the effect of pressure on the electronic structure of ge...Alexander Decker
The document analyzes the effect of pressure on the electronic structure of germanium nanocrystals using computational methods. It models germanium nanocrystals containing 8, 16, and 54 atoms using the large unit cell method within restricted Hartree-Fock theory. It calculates properties like cohesive energy, band gap, valence bandwidth, and bulk modulus under varying pressures. The results show that valence bandwidth decreases with pressure while cohesive energy and band gap vary depending on crystal size. Properties generally increase with crystal size. The maximum modeled pressure of 7.6 GPa agrees with the experimental germanium phase transition pressure.
Analytical model for the effect of pressure on the electronic structure of ge...Alexander Decker
This document analyzes the effect of pressure on the electronic structure of germanium nanocrystals using ab initio restricted Hartree-Fock theory. It finds that:
1) The valence bandwidth decreases with increasing pressure for all crystal sizes (8, 16, and 54 atoms) as atoms become more spaced out, reducing energy level splitting.
2) Band gap decreases with pressure for the 8- and 16-atom crystals but increases for the 54-atom crystal, representing a transition from semiconductor to insulator.
3) Cohesive energy decreases with tensile pressure for the 8-atom crystal but increases for the 16- and 54-atom crystals, possibly related to crystal shape changes from cubic
This document discusses parameterization of force fields for molecular mechanics simulations. It describes how parameters are determined for force fields like AMBER and CHARMM from experimental data and quantum calculations. The parameters include values for bonds, angles, dihedrals, van der Waals interactions, and electrostatic terms that describe the energy landscape for molecular simulations. Both bonding and nonbonding terms are considered in the parameterization process.
This editorial discusses the need for critical assessment of theoretical calculations of atomic structure and transition probabilities. It notes that while computational codes can generate extensive atomic data, the quality of this data is rarely verified. It suggests various methods for estimating uncertainties in theoretical atomic data calculations, including comparing length and velocity gauges, performing convergence studies, and internal benchmarking for large data sets. The goal is to develop strategies to evaluate theoretical atomic data and ensure its quality for applications.
This editorial discusses the need for critical assessment of theoretical calculations of atomic structure and transition probabilities. It notes that while computational codes can generate extensive atomic data, the quality of this data is rarely verified. It suggests various methods for estimating uncertainties in theoretical atomic data calculations, including comparing length and velocity gauges, performing convergence studies, and internal benchmarking for large data sets. The goal is to develop strategies to evaluate theoretical atomic data and ensure its quality for applications.
This editorial discusses the need for critical assessment of theoretical calculations of atomic structure and transition probabilities. It notes that while computational codes can generate extensive atomic data, the quality of this data is rarely verified. It suggests various methods for estimating uncertainties in theoretical atomic data calculations, including comparing length and velocity gauges, performing convergence studies, and internal benchmarking for large data sets. The goal is to develop strategies to evaluate theoretical atomic data and ensure its quality for applications.
This editorial discusses the need for critical assessment of theoretical calculations of atomic structure and transition probabilities. It notes that while computational codes can generate extensive atomic data, the quality of this data is rarely verified. It suggests various methods for estimating uncertainties in theoretical atomic data calculations, including comparing length and velocity gauges, performing convergence studies, and internal benchmarking for large data sets. The goal is to develop strategies to evaluate theoretical atomic data and ensure its quality for applications.
This document discusses the 3D numerical analysis of the electromagnetic field in a claw-pole alternator. It presents the development of a 3D finite element model of a claw-pole alternator to analyze the electromagnetic field distribution. The model considers the different magnetic properties of components like the rotor poles and stator core. The analysis uses the Newton-Raphson method to solve the nonlinear magnetic field equations over multiple iterations to obtain an accurate solution. The goal of the analysis is to study ways to improve alternator performance by reducing magnetic saturation in the claw poles.
Similar to Development of an artificial neural network constitutive model for aluminum 7075 alloy (20)
ESR spectroscopy in liquid food and beverages.pptxPRIYANKA PATEL
With increasing population, people need to rely on packaged food stuffs. Packaging of food materials requires the preservation of food. There are various methods for the treatment of food to preserve them and irradiation treatment of food is one of them. It is the most common and the most harmless method for the food preservation as it does not alter the necessary micronutrients of food materials. Although irradiated food doesn’t cause any harm to the human health but still the quality assessment of food is required to provide consumers with necessary information about the food. ESR spectroscopy is the most sophisticated way to investigate the quality of the food and the free radicals induced during the processing of the food. ESR spin trapping technique is useful for the detection of highly unstable radicals in the food. The antioxidant capability of liquid food and beverages in mainly performed by spin trapping technique.
Mending Clothing to Support Sustainable Fashion_CIMaR 2024.pdfSelcen Ozturkcan
Ozturkcan, S., Berndt, A., & Angelakis, A. (2024). Mending clothing to support sustainable fashion. Presented at the 31st Annual Conference by the Consortium for International Marketing Research (CIMaR), 10-13 Jun 2024, University of Gävle, Sweden.
Describing and Interpreting an Immersive Learning Case with the Immersion Cub...Leonel Morgado
Current descriptions of immersive learning cases are often difficult or impossible to compare. This is due to a myriad of different options on what details to include, which aspects are relevant, and on the descriptive approaches employed. Also, these aspects often combine very specific details with more general guidelines or indicate intents and rationales without clarifying their implementation. In this paper we provide a method to describe immersive learning cases that is structured to enable comparisons, yet flexible enough to allow researchers and practitioners to decide which aspects to include. This method leverages a taxonomy that classifies educational aspects at three levels (uses, practices, and strategies) and then utilizes two frameworks, the Immersive Learning Brain and the Immersion Cube, to enable a structured description and interpretation of immersive learning cases. The method is then demonstrated on a published immersive learning case on training for wind turbine maintenance using virtual reality. Applying the method results in a structured artifact, the Immersive Learning Case Sheet, that tags the case with its proximal uses, practices, and strategies, and refines the free text case description to ensure that matching details are included. This contribution is thus a case description method in support of future comparative research of immersive learning cases. We then discuss how the resulting description and interpretation can be leveraged to change immersion learning cases, by enriching them (considering low-effort changes or additions) or innovating (exploring more challenging avenues of transformation). The method holds significant promise to support better-grounded research in immersive learning.
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...Sérgio Sacani
Context. With a mass exceeding several 104 M⊙ and a rich and dense population of massive stars, supermassive young star clusters
represent the most massive star-forming environment that is dominated by the feedback from massive stars and gravitational interactions
among stars.
Aims. In this paper we present the Extended Westerlund 1 and 2 Open Clusters Survey (EWOCS) project, which aims to investigate
the influence of the starburst environment on the formation of stars and planets, and on the evolution of both low and high mass stars.
The primary targets of this project are Westerlund 1 and 2, the closest supermassive star clusters to the Sun.
Methods. The project is based primarily on recent observations conducted with the Chandra and JWST observatories. Specifically,
the Chandra survey of Westerlund 1 consists of 36 new ACIS-I observations, nearly co-pointed, for a total exposure time of 1 Msec.
Additionally, we included 8 archival Chandra/ACIS-S observations. This paper presents the resulting catalog of X-ray sources within
and around Westerlund 1. Sources were detected by combining various existing methods, and photon extraction and source validation
were carried out using the ACIS-Extract software.
Results. The EWOCS X-ray catalog comprises 5963 validated sources out of the 9420 initially provided to ACIS-Extract, reaching a
photon flux threshold of approximately 2 × 10−8 photons cm−2
s
−1
. The X-ray sources exhibit a highly concentrated spatial distribution,
with 1075 sources located within the central 1 arcmin. We have successfully detected X-ray emissions from 126 out of the 166 known
massive stars of the cluster, and we have collected over 71 000 photons from the magnetar CXO J164710.20-455217.
Current Ms word generated power point presentation covers major details about the micronuclei test. It's significance and assays to conduct it. It is used to detect the micronuclei formation inside the cells of nearly every multicellular organism. It's formation takes place during chromosomal sepration at metaphase.
Authoring a personal GPT for your research and practice: How we created the Q...Leonel Morgado
Thematic analysis in qualitative research is a time-consuming and systematic task, typically done using teams. Team members must ground their activities on common understandings of the major concepts underlying the thematic analysis, and define criteria for its development. However, conceptual misunderstandings, equivocations, and lack of adherence to criteria are challenges to the quality and speed of this process. Given the distributed and uncertain nature of this process, we wondered if the tasks in thematic analysis could be supported by readily available artificial intelligence chatbots. Our early efforts point to potential benefits: not just saving time in the coding process but better adherence to criteria and grounding, by increasing triangulation between humans and artificial intelligence. This tutorial will provide a description and demonstration of the process we followed, as two academic researchers, to develop a custom ChatGPT to assist with qualitative coding in the thematic data analysis process of immersive learning accounts in a survey of the academic literature: QUAL-E Immersive Learning Thematic Analysis Helper. In the hands-on time, participants will try out QUAL-E and develop their ideas for their own qualitative coding ChatGPT. Participants that have the paid ChatGPT Plus subscription can create a draft of their assistants. The organizers will provide course materials and slide deck that participants will be able to utilize to continue development of their custom GPT. The paid subscription to ChatGPT Plus is not required to participate in this workshop, just for trying out personal GPTs during it.
The cost of acquiring information by natural selectionCarl Bergstrom
This is a short talk that I gave at the Banff International Research Station workshop on Modeling and Theory in Population Biology. The idea is to try to understand how the burden of natural selection relates to the amount of information that selection puts into the genome.
It's based on the first part of this research paper:
The cost of information acquisition by natural selection
Ryan Seamus McGee, Olivia Kosterlitz, Artem Kaznatcheev, Benjamin Kerr, Carl T. Bergstrom
bioRxiv 2022.07.02.498577; doi: https://doi.org/10.1101/2022.07.02.498577
Or: Beyond linear.
Abstract: Equivariant neural networks are neural networks that incorporate symmetries. The nonlinear activation functions in these networks result in interesting nonlinear equivariant maps between simple representations, and motivate the key player of this talk: piecewise linear representation theory.
Disclaimer: No one is perfect, so please mind that there might be mistakes and typos.
dtubbenhauer@gmail.com
Corrected slides: dtubbenhauer.com/talks.html
The binding of cosmological structures by massless topological defectsSérgio Sacani
Assuming spherical symmetry and weak field, it is shown that if one solves the Poisson equation or the Einstein field
equations sourced by a topological defect, i.e. a singularity of a very specific form, the result is a localized gravitational
field capable of driving flat rotation (i.e. Keplerian circular orbits at a constant speed for all radii) of test masses on a thin
spherical shell without any underlying mass. Moreover, a large-scale structure which exploits this solution by assembling
concentrically a number of such topological defects can establish a flat stellar or galactic rotation curve, and can also deflect
light in the same manner as an equipotential (isothermal) sphere. Thus, the need for dark matter or modified gravity theory is
mitigated, at least in part.
The binding of cosmological structures by massless topological defects
Development of an artificial neural network constitutive model for aluminum 7075 alloy
1. Development of an Artificial Neural Network Constitutive Model
for Aluminum 7075 Alloy
Sanjeev Sen, Janet M. Twomey and Jamal Y. S. Ahmad
Department of Industrial and Manufacturing Engineering
Wichita State University, Wichita, KS 67260, U.S.A.
Abstract
Development of an accurate neuronal material model for aluminum alloys would facilitate better process control and
parameter optimization in the aircraft industry. This research primarily attempts to design a connectionist
constitutive model for aluminum 7075 alloy using multi-layer perceptrons. The backpropagation algorithm is used to
train the designed network whose performance is then validated using experimental data from a test set.
Keywords
Artificial Neural Networks (ANNs), Constitutive Modeling, Intelligent Systems, Connectionist Material Model
1. Introduction
Constitutive equations are mathematical relationships that describe the macroscopic response of a material to
applied stress under different combinations of strain ( ε), strain rate ( ε& ), and temperature (T). It is usually of the
form, ),,( Tf εεσ &= , where σ = flow stress. It is impossible to build a single set of equations to accurately
approximate the behavior of any metal. However we strive to formulate equations to predict material behavior under
a subset of conditions. The material behavior is thereby described; to be used in analysis, design of processes and
possibly in the development of the materials themselves. Hence constitutive equations are the mathematical
idealization of material behavior under a set of given conditions.
The typical approach to the development of constitutive models is parametric. The major drawback of parametric
modeling is the assumption of the model form, which relies heavily on the modeler’s expertise. As a consequence
the form taken by the final model may be too specific to generalize beyond new unseen data or too inaccurate to be
of use. System design using Artificial Neural Networks (ANNs) is a non-parametric modeling technique. The neural
network modeling technique can suffer from the same problems, being too specific or too inaccurate, but relies less
on the modeler’s expertise. Therefore they are more often better predictors. This paper reports preliminary results of
a neural network constitutive model of aluminum 7075 alloy. This research is a part of a larger body of work that
attempts to model high speed machining processes for the aircraft manufacturing industry.
2. ANN Constitutive Models and Mathematical Relationships
Mathematical constitutive models are rigid and may compute erroneous variable magnitudes during parameter range
shifts. Neural networks have contemporarily been used as alternatives to mathematical constitutive modeling. ANN
approaches have essentially been justified owing to the capabilities of neuron models to learn the non-linear
relationships between input and output parameters in the system. This feature of the ANN approach is in essence of
the complexity involved in quantifying the behavior of metals under varying external conditions. According to
Gaboussi et al [5], the inherent properties of ANNs in constitutive modeling as an alternative to classical rheological
modeling are; ability of adaptation, distributed memory, ability of generalization and a strongly parallel structure.
These properties significantly avoid sensitivity to noise and facilitate fast data processing. Fast data processing is
essential for its use in finite element analysis.
2.1 Parametric approach to constitutive modeling
The primary difficulty is to address nonlinear phenomena like elastoplasticity with cold work and viscoplasticity in
the material whose response is to be modeled. Various approaches have been researched in the past. A parametric
identification method as a starting procedure is described in Pernot and Lamarque [2]. The parameters are identified
from experimental time series that approximate displacement, strain and stress magnitudes for different quasi-static
loading conditions. In conformance to the specified relationships between stress states and the strain field, the
2. parameters that affect the model can be computed. Subsequently intrinsic formulations are investigated from
experimental data. The output from the model is compared with the experimental results marked apriori as the test
set. Thus the model performance is validated. Several iterations are required until the output converges to an
acceptable degree of agreement with the test set. The rheological framework of modeling constitutive equations
incorporates parametric identification and computation of intrinsic variables [1,4,6].
2.2 Non-parametric approach to constitutive modeling
As stated above, the ANN approach is classified as a non-parametric technique [3]. The convenience of the ANN
approach is that the understanding of complex nonlinear responses of the material is not necessary. The objective is
to design a connectionist model using data available as fitting parameters thereof. A rudimentary network is first
designed using experimental data. The identification of internal parameters of the network is then processed during
the training phase. The training of the network is an iterative process wherein the objective of the learning phase is
to get the designed network to simulate actual outputs of the physical system in correspondence to the inputs
presented. The training method or the network design is accordingly modified to obtain a close agreement between
network predicted and experimentally procured test magnitudes. The ANN methodology involves the computation
of weights in accordance to the backpropagation training in the learning phase. The advantages of the neural
approach to constitutive modeling [2] can be summarized as,
§ It allows to avoid apriori assumptions about the type of constitutive laws
§ It can solve the problem of constitutive law inversion
§ It can be used to control “rheological behavior” of an in-situ material
§ It can directly use experimental results in order to build a model
§ The training phase can be extended to improve the model
Neural based constitutive laws may be implicit or explicit depending on their mathematical description. Furukawa
and Yagawa [3] formulated an implicit relation, wherein ψis the implicit mapping in terms of the state space
method, ),( uxx ψ=& , x and u being state variables and control inputs. State space forms in various applications
have been learnt successfully by neural networks. This research would focus on the development of an implicit
constitutive model for Aluminum 7075. The most important advantage of the implicit model is that it can be
constructed from experimental input output data.
3. Material constitutive model for aluminum
Aluminum 7075 is among the most extensively used alloys for aircraft fuselage structures owing to its high strength-
to-density ratio. According to a generic observation, it is not realistic to express flow stress as a single function of
strain, strain rate and temperature that is applicable for all possible combinations of these parameters [11]. In
conventional constitutive theories, a mathematical model is constructed, primarily to represent the behavior of the
material at moderate ranges of temperature and strain rate. There also exists a class of constitutive relations that
model material flow properties and are pertinent to the microstructure of the metal subjected to external variables.
Work on metals has been documented from the viscoplastic perspective and extensive models have also been
designed for non-metals [10]. However, a robust model for aluminum and its alloys does not exist for high strain
rates and high ranges of temperature. The complexity of material responses for different segments of the parameter
range could not be incorporated by a static mathematical model. Thus the mathematical approach would require
defining multiple equations for different range domains. From the perspective of industrial processes involving
Aluminum 7075, a robust dynamic system that would predict the alloy behavior for the designated parameters over
any given process range is required.
4. Experimental data and parameter sensitivity
Experimental data of the dynamic impact properties of aluminum 7075 alloy were obtained using a Split
Compression Hopkinson Bar (SHPB). This method has been extensively used to procure material property data
under high strain rate magnitudes [7].
3. The experimental results were documented in a previous
study [8] as parameter responses for three different strain
rate sets, ε& =1300s-1
, 2400 s-1
and 3100 s-1
. The alloy
behavior at each of the strain rate magnitudes are
represented by figures 1, 2 and 3. The stress strain curves
are tabulated at four temperatures for each strain rate
response s et, T1 = 300 o
C, T2 = 200 o
C, T3 = 100 o
C and T4
= 25 o
C, represented in that order by the curves starting at
the bottom of each graph upwards. Flow stress (true
stress), represented on the y-axes of each of the 3 graphs is
a function of temperature (T), true strain (ε) and strain
rate (ε& ). After initial yielding, the flow stress exhibits a
steady increase with increase in true strain. This increase
is accompanied with different strain hardening rates in the continuum, a variable that has not been included in the
preliminary study. According to the sensitivity analysis, the effect of temperature on the flow stress is more pronounced
than that of the strain rate. This is further substantiated by an appreciable decrease in flow stress with a corresponding
increase in temperature for a constant strain rate. Also, the temperature sensitivity is independent of the strain rate.
Previous studies have also documented the evolution of the microstructure during deformation of the material specimen
[9, 12], as a consequence of external loading.
Various systems simulation processes have also been used to study aluminum alloy responses at different loading
parameters. The designed neural network model in this research would be trained to predict the output based on the
experimental data discussed in this section.
5. Network Architecture
The ANN model developed may be classified as a knowledge based constitutive relationship system. This model would
approximate the material behavior in a radically different way; moreover it will not be confined to the boundaries of a
computed mathematical model. The ANN model thereby designed would approximate the alloy response over the entire
continuum for the designated parameters.
The response characteristics in the form of true stress output magnitudes corresponding to various sets of temperature
(T), true strain (ε) and strain rate (ε& ) inputs, are learned by the neural network as representative functions to be
incorporated into the proposed neuron constitutive model. The global training method used is in the form of an
estimation of a function mapping (Rm
⇒Rn
) from samples of the function made available through experimental data
discussed in section 4. The characteristics of this function to be approximated vary throughout the input space.
The network design objective is therefore characterized as a function approximation state space modeled using a multi-
linear perceptron system with 3 input nodes each for temperature (T), true strain (ε) and strain rate (ε& ) in the input
layer and one node corresponding to the magnitude of true stress (σ ) in the output layer. The number of hidden layers is
experimentally determined. Among the designed networks tested, a four-layer network topology with 10 neurons in each
of the two hidden layers (3-10-10-1 network) exhibited the best performance.
0
150
300
450
600
750
900
1050
0 0.1 0.2 0.3 0.4
Figure 3 (strain rate = 3100/s) x=True Strain
y=TrueStress(MPa)
300o
C
200o
C
100o
C
25o
C
0
150
300
450
600
750
900
0 0.05 0.1 0.15 0.2 0.25
Figure 1(strain rate 1300/s) x=True Strain
y=TrueStress(MPa)
0
150
300
450
600
750
900
0 0.1 0.2 0.3
Figure 2 (strain rate=2400/s) x=True Strain
y=TrueStress(MPa)
300o
C
200o
C
100o
C
25o
C
300o
C
200o
C
100o
C
25o
C
4. 5.1 The ANN training algorithm
Algorithms like Classification And Regression Trees (CART), Multivariate Adaptive Regression Splines (MARS),
ID3, and the Hierarchical Mixtures of Experts (HME), are local approximation models [14], where the input space is
divided, at training time, into a hierarchy of regions where simple surfaces are fit to the local data [13].
The training method used to design the neuron constitutive model in this research is the backpropagation algorithm.
This supervised training algorithm incorporates an application of the chain rule for ordered partial derivatives to
calculate the sensitivity that a cost function has with respect to the internal states and weights of a network. The
training error is backward-passed to each internal node within the network, which is then used to calculate weight
gradients for that node. Learning therefore progresses by alternately propagating forward the activations and
propagating backward the instantaneous errors.
Root Mean Square Error (RMSE) is selected as the success quantifying metric for the backpropogation training of
the designed ANN constitutive model. The goal was to minimize RMSE over the entire range of the network
output. The backpropogation algorithm is used to train the network to approximate the alloy response data iteratively
minimizing the root mean square error of the computed output magnitudes.
The training set consisted of 562 input-output data pairs procured from experimental results discussed in section 4.
Training was stopped after the RMSE appeared to stabilize over several iterations. The final network was selected
by on the basis of the lowest RMSE on a separate validation or test set of data.
The validation set consisted of experimental data points over the same range as the training set. The validation set
had 209 input-output data pairs, which do not have occurrences in the training set.
6. Preliminary results and network performance
The performance of the networks was evaluated based on the output generated for stress (σ , MPa) versus the
experimental data on the validation set, which is shown in figures 5, 6 and 7. Each of the figures are plots of true
stress against true strain and is presented in the same format as the input data in section 4.
The four result curves in each figure correspond to temperatures 300, 200, 100 and 25 o
C. The curves attest that the
neural operator has accurately reproduced the nature of the curve and the ANN model has acquired the capability to
approximate the decrease in true stress with a corresponding decrease in temperature.
The identification of the correlation between the parameters by the designed network is significant, as it would
facilitate accurate network prediction of output magnitudes given any unlearnt data point outside the continuum of
the training set.
0
150
300
450
600
750
900
0 0.1 0.2 0.3 0.4
Figure 6 (strain rate 2400/s) x=True strain
y=Truestress(MPa)
0
150
300
450
600
750
900
1050
0 0.1 0.2 0.3 0.4
Figure 7 (strain rate 3100/s) x=True strain
y=Truestress(MPa)
300o
C
200o
C
100o
C
25o
C
300o
C
200o
C
100o
C
25o
C
__ Experimental data o ANN output
__ Experimental data o ANN output
5. Figure 8 represents the actual error of the output points generated by the ANN model corresponding to the
experimental test data presented to the network during the validation phase.
The initial points show that there was a slight incoherence, which may be attributed to partial data assimilation.
However, the error stabilizes and most data points are approximated to a perceptible degree of accuracy. The error
magnitudes are represented for all the 209 data points in the test set, maximum error being 6.72 MPa and the
minimum error being 0.0 MPa.
7. Concluding Remarks
This paper presents the preliminary results of current research in the development of a neuro-mimetic material
model. The credibility of artificial neural networks in approximating material response presented as parametric
trajectories has been established. Previous studies have documented ANN identification capabilities of non-linear
dynamical systems [10]. Many aviation materials are processed without information about their mechanical
behavior. Thus a tool that would accurately predict material behavior subject to machining parameters would
facilitate greater process control in machining aviation alloys. A connectionist model may therefore be designed for
any industrial material, based on experimental data and mathematical constitutive laws available. The ANN model
would therefore be an evolutionary tool in the field of material constitutive modeling.
8. Anticipated results of ongoing research
Presently, a training procedure using smaller training sets over larger epochs is being explored. The research is also
investigating another approach wherein multiple networks approximate the behavior of the alloy in a discretized
input parameter domain, one network for each parameter range. The validity of the ANN predicted results outside
the range of the training set are being experimentally verified. The results from the final connectionist model would
be more accurate than corresponding mathematical and finite element models for the system.
-8
-6
-4
-2
0
2
4
1 26 51 76 101 126 151 176 201
Figure 8 x=Serial number of experimental data point
y=Networkoutputerror
0
150
300
450
600
750
900
0 0.05 0.1 0.15 0.2 0.25
Figure 5 (Strain rate = 1300/s) x=True strain
y=truestress(MPa)
Experimental data ANN output
300o
C
200o
C
100o
C
25o
C
6. References
1. Ghaboussi, J., Garret, J.H., Jr., and Wu, X. (1991). Knowledge-Based Modeling of Material Behavior with
Neural Networks. Journal of Engineering Mechanics Division, ASCE, 117 (1), 132-153
2. Pernot, S., & Lamarque, H. C. (1997). Ecole Nationale des Travaux Publics de l’Etat, France.
3. Furukawa, T., Okuda, H., & Yagawa, G. (1996). Implicit constitutive modeling using neural networks. 19th
International Conference of Theoretical and Applied Mechanics, Kyoto, 25-32 August, 531 pp
4. Ghaboussi, J., Lade, P. V., & Sidarta, D. E. (1994). Neural Networks Based Modeling in Geomechanics.
Proceedings of the 8th
International Conference on Computer Methods and Advances in Geomechanics,
Morgantown, West Virginia.
5. Ghaboussi, J., Pecknold, D. A., Zhang, M., & Haj-Ali, R.M. (1998). Autoprogressive Training of Neural
Network Constitutive Models.
6. Ghaboussi, J., & Sidarta, D. E. (1998). New Nested Adaptive Neural Networks (NANN) for Constitutive
Modeling. Computers and Geotechnics, 22(1), 29-52.
7. Ghosh, A. K., Gandhi, G., in: H. J. Mc Queen et al. (Eds.), Strength of Metals and Alloys (ICSMA7), Pergamon
Press, Oxford, 1985, pp 2065.
8. Lee, S. W., Sue, C. W., & Lin, W. J. C. (1998). The strain rate and temperature dependence of the dynamic
impact properties of 7075 aluminum alloy. Journal of Materials Processing Technology. 100, 116-122.
9. T Sheppard and D S Wright Determination of flow stress: Part 1 Constitutive equation for aluminum alloys at
elevated temperatures. The Metals Society. 1978.
10. Chen, S., & Billings, S. A. (1992). Neural networks for nonlinear dynamic system modeling and identification.
International journal of control, 56(4), 319-346.
11. A. K. Miller, An inelastic constitutive model for monotonic, cyclic, and creep deformation. Part -I Equations
development and analytical procedures; Part –II Application to type 304 stainless steel, Journal of Engineering
Materials and Technology, 98H (1976) 97-113.
12. T. Sheppard and D. S. Wright. Determination of flow stress: Part 1 Constitutive equation for aluminum alloys at
elevated temperatures. The Metals Society. 1978.
13. Fausett L., 1994, Fundamentals to Neural Networks: Architectures, Algorithms, and Applications. Prentice
Hall, New Jersey.
14. Gibbon, J. D., 1997, Nonparametric Method for Quantitative Analysis, American Sciences Press, Columbus,
Ohio.